Lead-acid and Lithium-ion Batteries for the Chinese Electric Bike Market and Implications on Future Technology Advancement

Jonathan X. Weinerta,*, Andrew F. Burkea, Xuezhe Weic

a

Institute of Transportation Studies University of California Davis 1 Shields Ave, Davis CA 95616, USA b Institute of Transportation Studies University of California Davis 1 Shields Ave, Davis CA 95616, USA c Tongji University, School of Automotive Studies 4800 Cao An Lu, Shanghai 201804, China *Corresponding author: jxweinert@ucdavis.edu, 86-138-1813-1364 or 001-530-752-1599 (phone), 1 Shields Ave, Institute of Transportation Studies, Davis CA 95616

Abstract China has been experiencing a rapid increase in battery-powered personal transportation since the late 90s due to the strong growth of the electric bike and scooter (i.e. e-bike) market. Annual sales in China reached 17 million bikes/yr in 2006. E-bike growth has been in part due to improvements in rechargeable valve-regulated lead acid (VRLA) battery technology, the primary battery type for e-bikes. Further improvements in technology and a transition from VRLA to lithium-ion (Li-ion) batteries will impact the future market growth of this transportation mode in China and abroad. Battery performance and cost for these two types are compared to assess the feasibility of a shift from VRLA to Li-ion battery e-bikes. The requirements for batteries used in e-bikes are assessed. A widespread shift from VRLA to Li-ion batteries seems improbable in the near future for the mass market given the cost premium relative to the performance advantages of Li-ion batteries. As both battery technologies gain more realworld use in e-bike applications, both will improve. Cell variability is a key problematic

area to be addressed with VRLA technology. For Li-ion technology, safety and cost are the key problem areas which are being addressed through the use of new cathode materials. Keywords Battery, electric bike, electric scooter, valve-regulated lead-acid, lithium-ion

Acronym List VRLA- valve-regulated lead acid, SSEB- scooter style electric bike, BSEB- bicycle style electric bike, AGM- absorptive glass mat, FLA- flooded lead acid, LAB- lead-acid battery, SLI starting, lighting and ignition

1. Introduction The Chinese electric bike market has expanded more rapidly than any other vehicle mode in the last seven years, from nearly 40,000 in 1998, to an estimated 15-18 million in 2006 [1, 2]. Electric bikes are a category of vehicles in China that includes two-wheel bicycles propelled by human pedalling supplemented by electrical power from a storage battery, and low-speed scooters propelled almost solely by electricity (usually with perfunctory pedals to satisfy legal definitions). These two types (referred collectively henceforth as e-bikes), are shown in Figure 1, include bicycle-style e-bike (BSEB), which resembles a regular bicycle, and scooter-style e-bike (SSEB), which is typically heavier and bigger. These vehicles have become a popular transportation mode for Chinese consumers because they provide an inexpensive and convenient form of private mobility

and are thus an attractive alternative to public transit or regular bicycling. National and many local governments promote them due to their low energy consumption and zero tail-pipe emissions, especially important in Chinas congested urban areas. E-bikes are gaining an increasing share of two-wheeled transportation throughout China, and in some cities like Chengdu and Suzhou, have even surpassed bicycle mode share proportion. Electric bikes have been by far the most successful battery electric vehicle application in history with estimated cumulative production of ~30 million by 2007 [3]. At the heart of e-bike technology is the rechargeable battery. The core rechargeable battery technology used in e-bikes is valve-regulated lead-acid (VRLA), or sealed, and lithium-ion (Li-ion). Advances in VRLA batteries and rising gasoline prices over the past decade have made e-bikes increasingly competitive with gasoline scooters in price and performance [4]. E-bikes using VRLA achieve low cost ($150-300) and adequate range (30-70km per 8 hr charge). The power system characteristics of e-bikes are shown in Table 1. Because most e-bikes use either VRLA or Li-ion, this analysis will focus on these two battery types. 1.1 Motivation Urbanization and income levels are both rising rapidly in China. As a result, the low and middle classes are demanding faster, more comfortable transportation options. Survey data from three major cities in China indicate that todays bicycle users (450 million) will most likely purchase an e-bike as their next mode of transport [5]. This mode shift choice has far-reaching impacts around the world in terms of battery technology development and the future market of electric two-wheelers.

This market development has important implications on the environment and public health. E-bikes are an extremely energy efficient (~50 mile/kWh) mode of personal transportation with zero tailpipe emissions. Life-cycle carbon emissions per km travelled of an e-bike are roughly one fifth that of a gasoline-fuelled car [6]. E-bikes are also having positive effects in cities battling poor air quality by displacing gasolinepowered scooters.

1.2 Methodology The analysis relies on literature and data from surveying a variety of companies involved in battery production for e-bikes. The authors visited several battery factories making both Pb-acid and Li-ion batteries. Batteries from some of these manufacturers have been laboratory-tested.

2. Transportation Battery Applications and Requirements Batteries are used for a wide range of applications including consumer electronics, energy, industrial, and transportation. Batteries for transportation applications have much different requirements than most other applications. They are used in three different modes: motive power, auxiliary power, and traction. Motive power batteries are used to drive automobiles, scooters, and bicycles and thus require high specific energy (Wh/kg) to achieve adequate range. Deep-cycling capability is necessary since it is common for batteries to be discharged to 10-20% SOC.

Cost is a driving factor because the battery pack size can be quite large (1-2 kWh for large electric scooters). Auxiliary power batteries are used in automobiles and motorcycles predominantly for starting, lighting, and ignition (SLI). Power is valued more than energy density and deep discharge capability because SLI batteries are primarily are used to provide high bursts of power output to start an engine (~1-5 kW) and rarely discharged more than 20%. Traction batteries used in fork-lifts and underground mining cars experience heavy-duty operation and thus require high abuse-resistance. These applications typically use flooded Pb-acid (FLA). Table 2 summarizes these battery applications and their requirements.

3. The Battery Industry In China The total Chinese battery market in China was valued at $12.4 billion in 2006, 35% of which is for rechargeable Pb-acid type. Estimates on the production volume capacity (kWh) of Pb-acid batteries range from 35-67 million kWh/yr, produced by more than 2,000 companies [3, 7]. 300 of these companies specialize in e-bike batteries with an estimated annual production between of 3.5 and 9 million kWh/yr in 2005. Calculations based on the annual e-bike sales in 2006 and assumed after-market sales to the existing ebike population reveal a much higher annual production between 15 and 20 million kWh/yr [8]. Figure 2 shows the proportions of different battery types in China.

VRLA batteries were first introduced into UPS applications in America and Europe in the 70s because of their low maintenance requirements and high reliability over traditional flooded lead-acid [9]. The rapid growth in telecommunication and computer networks throughout the world during the 80s created a huge market for this battery type. The VRLA industry finally spread to China in response to their telecommunications boom of the 90s [10]. Prior to the 90s, the Chinese battery industry produced mainly flooded Pb-acid batteries for agriculture and transport (e.g. trucks, train infrastructure). Between 1990 and 1996, sales of VRLA batteries grew from 60,000 to 730,000 kWh, primarily for telecommunications applications. In the late 90s, production of small VRLA and flooded SLI batteries grew in response to the growing automobile, gasoline scooter, and electric bike markets [11]. One of the main problems with Chinas Pb-acid battery industry is that it is difficult for government to regulate production, quality, and environmental impacts. This is in part due to the large number of relatively small manufacturers spread throughout the country. This high industry decentralization results in lower quality batteries entering the market and batteries containing toxic performance enhancing materials such as Cadmium, and lead waste issues. In 2006, 23% of the e-bike battery companies inspected did not pass the minimum quality standards set by the national inspection bureau [7]. It is expected that considerable consolidation within the industry will occur, as occurred in the European battery industry during the 90s [11, 12]. The advanced battery market in China makes up 15% of the total market, which includes batteries using lithium or nickel compounds. These companies primarily produce batteries for consumer electronics applications used throughout the world. The

first Li-ion battery was commercialized by Sony in 1991 in Japan for use in consumer electronics. Few LAB manufacturers in China are making advanced batteries. From one manufacturers perspective, Li-ion batteries are still dangerous, costly, and the market for LABs is still large and expanding.

3.1 VRLA Production Most of the worlds small VRLAs (<25Ah) are manufactured in Asia and exported around the world due to low labor costs, land cost, and loose environmental standards [11]. The process for making large modules is roughly the same as making small modules. Manufacturing is labour intensive yet exhibits low profit margins. Battery quality can be considerably different among manufacturers and is a key distinguishing factor between top brands from the hundreds of smaller competitors. Key differences from company to company are linked to differences in materials (i.e. alloy plate formula, electrolyte formula, AGM material, etc.) and manufacturing (i.e. dust control, quality inspection stations, etc) [13].

3.2 Li-ion Production Li-ion batteries, whether for electric vehicles, electric bikes, and consumer electronics, are all produced using similar processes [14]. Hence, a single manufacturer can produce battery sizes for a wide range of applications, from portable consumer electronics to EVs [15]. Li-ion batteries can be designed for high power or high energy depending on cell size, thickness of the electrode, and relative quantities of material used [14]. High power

cells are generally smaller in order to dissipate the higher heat load. Both types use the same current collectors and separators. Lithium resources are abundant in China. As of 2000, they were the 2nd largest producer of Lithium in the world and in 2004 produced 18,000 metric tons [16, 17].

4. Batteries For E-Bikes This section describes VRLA and Li-ion batteries for use in e-bikes and identifies the most important battery characteristics in this application. Based on the thriving market, todays batteries appear to satisfy the cost, range, weight, and other requirements demanded by e-bike users.

4.1 VRLA In 2005, it is estimated that 95% of e-bikes produced in China used VRLA; the rest use Li-ion, Ni-MH, or NiCd though the majority of these are exported [1]. VRLA battery packs consist of three to four 12V modules (12, 14 or 20Ah capacity) for a total voltage of 36 or 48V and energy capacity of 0.4-1 kWh. VRLA for e-bikes differ from SLI VRLAs used in automotive applications in that they are able to be deep-cycled. E-bike batteries are typically of the absorptive glass mat (AGM) type, meaning they use an absorbed sulphuric acid electrolyte in a porous separator, as opposed to a gelled silica/acid separator in Gel-type VRLAs. Whereas standard SLI automotive batteries are typically only discharged 10-15%, deep-cycle batteries for motive applications like ebikes are discharged 80-90% [18]. Battery makers claim the key distinguishing factor of 8

their batteries is lifetime and stability (i.e. mean time before failure). Most domestic manufacturers do not report defect rate of their products, but one study by a battery manufacturer reports a 3-9% defect rate of e-bike batteries from three domestic manufacturers.

4.2 Lithium ion Li-ion battery packs for e-bikes range from 24V-37 V with capacity of 5-60 Ah. The market for Li-ion e-bikes in China is still small. In Japan and Europe however, Li-ion and Ni-MH are the dominant battery type, though annual e-bike sales in these regions are two orders of magnitude lower than in China.

4.3 E-bike Battery Requirements VRLA is the current dominant technology in e-bikes. Li-ion and Ni-MH battery manufacturers are trying to tap into this large growing market. Some Li-ion battery companies are expecting 100% growth in sales in the next year and predict the market for Li-ion battery e-bikes will grow to 20% of total annual e-bike sales in the next 5 years. Ultimately, the battery type that succeeds will depend on its performance relative to the alternative based on the several key criteria, described below. These criteria are compared for VRLA and Li-ion batteries in Section 5. Cost: battery cost is likely the most critical factor in battery choice, as evidenced by the market dominance of VRLA. Despite the significant advantages in energy density and

lifetime of Li-ion, VRLA is much lower cost. The emphasis on cost may change as average income increases throughout China. Cycle life: lifetime of the battery is critical because it affects users long-term operating cost. E-bike length of ownership can last 3-5 years depending on use. However, most users find they need to replace their battery after 1-2 years due to serious performance degradation [19]. Weight: Vehicle range is one of the most critical metrics for e-bike users due to the long recharge times. Range depends on stored energy capacity, which for a given specific energy (Wh/kg), determines battery weight. Weight for VRLA e-bike batteries typically range from 12kg for BSEB to 26kg for large SSEB. There are practical battery weight limitations based on the users physical strength, since some users require removing the battery from the bike to recharge it in their apartment/home/office. Many users however often roll the entire e-bike into their house/apartment if there is an elevator or find a convenient place to recharge on ground level. In terms of practical limitations on consumer for demand range, surveys of e-bike users in three medium to large-sized cities show that average commute distance is 9.3 km/day [4]. However, there are high-range ebikes on the market that can achieve range of up to 80 km/charge. Charging safety: charging for VRLAs is considerably more flexible and tolerant to improper recharging than Li-ion batteries in terms of risk of damage to self and property. As evidenced by the worldwide Sony battery recall of 2006, Li-ion batteries still entail risk, which is amplified as cell size increases. Temperature effects: e-bike batteries are used over a wide range of temperatures from winter lows of -40 C in Chinas northeast to summer highs of +40 C in the southwest. 10

A batteries performance at extreme temperatures will affect range and lifetime and is thus an important factor.

4.3.1 Other Factors Volume: volume is likely a secondary factor since the weight constraint of a battery limits energy capacity before volume is constrained. Batteries for SSEBs are usually stored in the floorboard underneath the feet, or for BSEBs along one of the frames crossbars. The largest battery pack in a SSEB is roughly 9.3L. Extra volume through smaller battery size may be valued slightly for extra storage space. Speed: top speed is determined by battery power density and motor size. The power density of VRLA 230W/kg is more than sufficient to meet the 350W peak motor power limit of e-bikes. While national e-bike standards limit top speed to 20km/h, most BSEBs can reach 25-30km/h, and high-power SSEB can reach 35-40. Charge time: Since most people charge their battery overnight when electricity is cheapest, the maximum acceptable charge time is likely 8hrs. Full battery capacity should be restored to full charge by an 8-hour charge regime using 220V AC.

5. E-Bike Battery Performance And Price Advances in VRLA technology over the past decade have made e-bikes affordable, efficient, and practical [4]. Li-ion technology has also improved to a point such that Liion e-bikes are now marketed in China, in addition to being exported throughout the

11

world. The technical performance and price of VRLA and Li-ion batteries from Chinese manufacturers are compared in this section.

5.1 VRLA Battery Performance and Price The key performance characteristics and price of VRLA (AGM type) batteries from several manufacturers for two popular e-bike battery module sizes (20 and 12 Ah) are compared in Table 1. VRLA costs for 12V-12Ah modules from three Chinese and one Japanese brand are compared in Table 3. Note that the batteries tested are specifically designed for motive power, not SLI applications, which have different characteristics when deep-discharged. To verify performance, we obtained 12V-12Ah modules from 4 large e-bike battery suppliers and measured their performance using an Arbin BT2043 battery testing device. Current and power levels for these tests were chosen based on the typical demands of an electric bike. Table 4 shows the results of the tests. The discharge characteristics are given in Ah, Wh/kg, W/kg at 9.6V. The results from the tests exceed the manufacturers stated claims on energy density and are considered quite good for VRLAs of such small cell size.

5.1.1 Cycle life Since cycle-life testing requires over a year, the authors relied on data provided by manufacturers and warranty data. Manufacturers report cycle life between 400-550 cycles, though independent testing of 4 brands by an anonymous manufacturer revealed 12

cycle life of 300-400 cycles. This corresponds with typical 1 to 1.5 year warranties provided by most e-bike manufacturers.

5.1.2 Defect Rate The industry average defect ratio for e-bike batteries is 5% while only 0.10% for other types of LABs [20]. The main reason for this large difference is the extreme variation in charging and discharging experienced in e-bikes compared to other applications. There was a noticeable difference in the defect rate of foreign brands compared to Chinese brand LABs. According to interviews with one battery company, improving battery lifetime and stability is the key area of research.

5.2 Lithium-ion Performance and Price Li-ion battery performance and price from various Chinese and international manufactures is compared in Table 5. Prices range from $510-760/kWh. Due to the limited amount of companies making Li-ion e-bike batteries, price data is presented from only three companies. Data from another Chinese Li-ion battery manufacturer quotes cost between $300-600/kWh (retail price is not provided) [21]. The stated cycle life of Li-ion batteries from 3 manufacturers is 600-800 cycles. The actual warranty on their batteries is 2 years.

6. Battery Transitions In The E-Bike Market

13

The transition from VRLA to Li-ion batteries in e-bikes is progressing in China, based on interviews with Li-ion battery companies. The pace and extent of this transition is still uncertain, since the e-bike market is currently very cost-conscious. The following section uses the battery performance and cost data and battery choice criteria from the previous sections to compare e-bikes using VRLA versus Li-ion.

6.1 Comparison of key factors for VRLA and Li-ion: The characteristics of VRLA and Li-ion batteries are compared in Table 6. The batteries are sized for an average 48V SSEB with 60km range (0.90 kWh) and 350W motor. This type of e-bike was chosen since it is a popular model for a 3-person family. It sets a practical upper bound to maximum battery size in an e-bike, and is comparable in performance to a 50cc gasoline scooter. Battery characteristics assumptions are also listed in the table. An e-bike energy consumption of 0.014 kWh/km, and average user travel distance of 15 km/day was assumed in making the battery comparisons. The effect of a smaller battery weight on energy consumption was neglected. These results suggest that the cost differential between the battery types dominates all other factors. The added lifetime from the more durable Ni-MH and Li-ion is likely not valued very high si.3nce the life of an e-bike is not much greater than 3-4 years. The 18 kg mass difference between Pb-acid and Li-ion, however, is significant since a 26 kg battery is likely unmanageable for the majority of e-bike users. If those users only option to recharge is to carry the battery indoors, they may be inclined to use Li-ion.

14

6.2 Japan and Europe After China, the next largest e-bike market is Japan- with annual sales of 270,000 bikes/yr in 2006 and 13% average annual growth since 2000 [22]. In Europe, the market is estimated at 190,000 bikes/yr in 2006 [23]. Electric bikes in these markets are different from Chinese e-bikes in that these bikes are typically the pedal-assist type or pedelec. The user typically pedals but is assisted by a small electric motor when extra power is desired (e.g. acceleration, uphill climbs). Most pedelec e-bikes use Ni-MH or Li-ion batteries. Battery capacity ranges from 0.2-0.6 kWh, motor size ranges from 150-250W, and prices range from $700-2,000.

7. E-bike Market Growth and Battery Technology Advancement The growing e-bike market will necessarily lead to further advancements in battery technology and a gradual transition to more advanced battery technologies. In turn, this battery advancement will expand the market for e-bikes in China and throughout the world, especially in developing countries with high two-wheeler use. This section describes the importance of technology learning to the advancement of battery technology, and the key areas where this learning is most important. The principal reason for anticipated improvements in battery performance and cost is due to technological learning effects. There are three categories of learning

15

associated with technology development: research and development (R&D), manufacturing, and in-service use. The e-bike battery market is accelerating learning in all three categories. Li-ion battery production, whether for electronics or e-bikes, achieves learning in the first two categories because the materials and manufacturing process for large and small cells are similar [24]. Only e-bike battery production, however, will drive the operational learning progress for large format battery cell technology. The key areas of technology improvements for which this type of learning will have the most impact are: safe charging and discharging, cell degradation over time, operation in extreme environments (low and high temperatures), and cell variability within a battery pack and its effects on lifetime. Cell variability is a key issue with VRLA cells. Safety and cost are the key issues with Li-ion cells. These issues are explained in the following sub-sections.

7.1 Cell Variability VRLA batteries exhibit considerable scatter in performance (i.e. no two modules have exactly the same electrical characteristics). This results from slight variations in the properties of materials and the electrodes used to assemble the cells due to the imprecise, labor-intensive manufacturing process [25]. When connecting several modules in series as in the case of a 36V (3 module) or 48V (4 module) e-bike battery pack, there is often significant variability in the module voltage. This causes accelerated aging since the weakest module of the pack ages more rapidly [25].

16

7.2 Safety For Li-ion batteries, safety risks such as battery overheating, combustion, and explosive disassembly increase with the amount of energy contained within the cell/battery pack. LiCoO2 is commonly used material for small cell Li-ion batteries, but considered unsafe for large-format batteries [21]. New cathode materials such as LiFePO4 are being introduced into Li-ion batteries for e-bikes, resulting in significantly safety improvements [26]. Hot-box heating and overcharge testing reveal safety advantages of LiFePO4 over both LiMn2O4 and LiMn2O4 [27].

7.3 Cost Li-ion battery technology is still relatively new (12 years) so there are potentially many opportunities for cost reductions. Material substitution could make a large impact since 75% of the total battery cost is due to materials [14]. Research and development efforts are focused on using more inexpensive and chemically stable materials such as LiFePO4 and Li(Ni1/3Co1/3Mn1/3)O2 for the cathode. Table 7 presents the cost, energy density, and cycle life differences between the commonly used LiCoO2 cathode and these two alternative materials. For LiFePO4, energy density is sacrificed for lower cost and longer life, along with the safety advantages mentioned above.

8. Conclusions There has been a rapid transition to electric bikes and scooters in China with the market reaching nearly 16 million/yr in 2006. This e-bike growth has been in part due to 17

improvements in rechargeable valve-regulated lead acid (VRLA) battery technology in China. Further growth in the market and a transition from VRLA to lithium-ion batteries will in turn lead to greater improvements in performance and cost. VRLA and Li-ion battery technology for e-bikes has been assessed. For VRLA, a specific energy of 34Wh/kg and a cost of $88/kWh were determined for a number of international brands. Li-ion batteries in China on average have specific energy of 106Wh/kg and cost of $590/kWh. A widespread shift from VRLA to Li-ion batteries seems improbable for the mass market given the cost premium relative to the performance advantages of Li-ion batteries. As both battery technologies gain more real-world use in e-bike applications, both will improve. Cell variability is a key problematic area to be addressed with VRLA technology. For Li-ion technology, safety and cost are the key problem areas, which are already being addressed through the use of new materials such as LiFePO4.

9. Acknowledgements The authors would like to thank Joan Ogden, Dan Sperling, and Jack Johnston for their thoughtful comments and review of this work. We would also like to acknowledge the valuable assistance given by Yang Wei and Bi Lu of Tongji University in conducting interviews, the support of Prof Ma Jianxin and Tongji University College of Automotive Studies. We would also like to acknowledge Jennifer Guan, Wang Jiqiang, Ritar Battery, Lantian Double-cycle, and Xingheng for help in acquiring data.

8. References

18

1 Jamerson, F. Electric Bicycle World Report 2007 (8th Edition) www.ebwr.com. 2 Economic Research Force of the China State Council Development Center. Development Strategy of LEV in China Research Report (Beijing, 2006). 3 Wang, J. Brief Overview of Chinese Battery Industry & Market. China International Battery Forum, Beijing, China, 2006. 4 Weinert, J., Ma, C. and Cherry, C. The Transition to Electric Bikes in China: History and Key Factors for Rapid Growth. Journal of Transportation Special Issue: Motorization in Asia, 2007, 34(3), pp.26. 5 Weinert, J., Ma, C., Yang, X. and Cherry, C. The Transition to Electric Bikes in China and its Effect on Travel, Transit Use, and Safety. Transportation Research Board, Washington D.C., 2006. 6 Cherry, C. Implications of Electric Bicycle Use in China: Analysis of Costs and Benefits. UC Berkeley Center for Future Urban Transport-Volvo Summer Workshop, Berkeley CA, 2006. 7 Center for Battery Quality Inspection Quality of Lead-acid Batteries of Electric-Assist Bicycle Analysis Report. Beijing 2006. 8 Assuming average e-bike battery capacity is 0.53-0.67 kWh, 16 million new sales in 2006, of the estimated 20 million existing e-bikes, 40% replace their battery each year, 60% replace it every other year. . 9 Fouache, S., Chabrol, A., Fossati, G., Bassini, M., Sainz, M.J. and Atkins, L. Effect of calcium, tin and silver contents in the positive grids of automotive batteries with respect to the grid manufacturing process. Journal of Power Sources, 1999, 78(1-2), 1222. 10 Wang, Z. Manufacture and application of valve-regulated lead/acid batteries in China. Journal of Power Sources, 1998, 73(1), 93-97. 11 Eckfeld, S., Manders, J.E. and Stevenson, M.W. The Asian battery market--a decade of change. Journal of Power Sources, 2003, 116(1-2), 32-39. 12 Razelli, E. Prospects for lead-acid batteries in the new millenium. Journal of Power Sources, 2003, 116(1-2), 2-3. 13 Ritar Power Co. Personal communication, Shanghai, 2006. 14 Gaines, L. and Cuenca, R. Costs of Lithium-Ion Batteries for Vehicles. Center for Transportation Research, Energy Systems Division, Argonne National Laboratory, 2000. 15 Broussely, M. Recent developments on lithium ion batteries at SAFT. Journal of Power Sources, 1999, 81-82, 140-143. 16 Tse, P.-K. The Mineral Industry of China. China-2004 (United States Geological Survey, 2004). 17 Ober, J. Lithium. U.S. Geological Statistics Yearbook (United States Geological Survey, 1999). 18 Moseley, P.T. High rate partial-state-of-charge operation of VRLA batteries. Journal of Power Sources, 2004, 127(1-2), 27-32. 19 Center for Electric Bicycle Quality Inspection Quality of Electric Bicycle Analysis Report. Beijing 2006. 20 Comparitive study of battery performance from large e-bike battery suppliers, conducted by one battery manufacturer, 2006. 21 Huang, S.-y. and Xiao, S. High Safety and Low Cost Polymer Li-ion Power Battery. China International Battery Forum, Beijing, China, 2006.

19

22 Suzuki, K. Advanced Technologies Driving Pedelec Market & Excellent Performance and Safety by Li-ion Battery for Pedelec. Light Electric Vehicle Conference, Taipei, Taiwan, 2007. 23 Neupert, H. Proposal to accelerate market size and quality of e-bikes/pedelecs in Europe. Light Electric Vehicle Conference, Taipei, Taiwan, 2007. 24 Anderman, M. Brief Assessment of Improvements in EV Battery Technology since the BTAP June 2000 Report. Sacramento, CA, 2003. 25 Rossinot, E., Lefrou, C. and Cun, J.P. A study of the scattering of valveregulated lead acid battery characteristics. Journal of Power Sources, 2003, 114(1), 160169. 26 Wu, D. Lithium-ion Phosphate Batteries: Enviro-friendly Technology Offered by Pihsiang Energy Technologies. Light Electric Vehicle Conference: Battery Safety, Taipei, Taiwan, 2007. 27 Yang, M.-h. Outlook of Future Li-ion Battery Chemistries for Safety Improvements. Light Electric Vehicle Conference: Battery Safety, Taipei, Taiwan, 2007. 28 Dell, R.M. and Rand, D.A.J. Understanding Batteries. (The Royal Society of Chemistry, Cambridge, 2001). 29 Bullock, K.R. Lead-acid battery research and development--a vital key to winning new business. Journal of Power Sources, 2003, 116(1-2), 8-13. 30 Chan, C.C. and Chau, K.T. Modern Electric Vehicle Technology. (Oxford University Press, Oxford, 2001).

Figure Captions:
Figure 1: Bicycle style and scooter style electric bikes Figure 2: The China Battery Market by Battery Type

Table 1: Electric two-wheeler power system characteristics Specifications Bicyclestyle ebike 0.4-0.6 15 36 3 18 0.24 80 Scooterstyle ebike 0.8-1.0 20-30 48 4 24 .50-1 80

Total battery pack capacity (kWh) Maximum current Voltage Modules/pack (typical) Cells in series Peak motor power (kW) Maximum depth of discharge (%)

Table 2: Battery Applications [28]

20

Application Electronics Energy Industrial Transportation

Function Portable power Remote-area power supply Back-up power Motive Power (hybrid) Motive Power (battery only) Auxiliary power (SLI) Traction

Battery Size 10 -10 Wh 103-105Wh [29] 102-103 Wh 102-103 Wh 102 Wh, 103 Wh
1 2

Technology Li-ion, Ni-MH FLA, VRLA

Requirements Low weight & volume, high energy Low maintenance, high reliability, long life

Ni-MH, Li-ion VRLA, Ni-MH, Li-ion FLA, VRLA FLA, VRLA

High specific power High specific energy, low cost low cost, high reliability Abuse tolerant, long life, low cost

Table 3: 20 and 12Ah VRLA module characteristics of various manufacturers1 Manufacturer Capacity (Ah) (2hr)2 12 12 10 12 10 12 Weight (kg) 4.4 4.1 4.1 3.8 4.1 4.2 Volume (L) 1.39 1.39 1.39 1.39 1.39 1.39 Specific Energy (Wh/kg) 33 35 29 38 29 34 33 20 20 20 20 20 7.2 10 6.6 7.0 6.8 2.37 3.63 2.30 2.31 2.40 33 24 36 34 35 33 Energy Density (Wh/L) 104 104 86 104 86 104 97 101 66 104 104 100 95 $88 Cost ($/kWh) 86.4 80.5 81.9 104.3

Ritar Tian Neng Chaowei Panasonic Sunbright Huafu AVERAGE Ritar Chaowei Panasonic Sunbright Huafu AVERAGE
1.

Information obtained from company websites. Price data is the purchase price from a retailer. A 2 hr rate is a commonly used metric for testing battery capacity. It represents the discharge rate used to completely discharging the battery in 2 hrs

2.

Table 4: Performance of 12V-12Ah VRLA Battery Modules from 4 Battery Manufacturers (C/2.4 discharge rate)

21

Company 1 2 3 4 Average

Mass (kg) 4.24 4.05 4.27 4.00 4.14

Capacity (Ah) 12.0 12.2 12.1 11.5 12.0

Specific Energy (Wh /kg) 34.2 36.8 34.3 35.0 35.1

Resistance (m) 20 22 27 30 25

Max power at 9.6V (W/kg) 272 258 200 192 231

Table 5: Characteristics of Li-ion modules from various manufacturers Manufacturer

Capacity (Ah) (2hr) Weight (kg) Volume (L)

Specific energy (Wh/kg) 63 68

Energy density (Wh/L) 128 173 151

Power density (W/kg) 1,261 1,805 1,533 111 200 unavail. unavail. unavail. unavail. unavail. unavail.

Price ($/kWh)

Xingheng - high power Xingheng - high power AVERAGE- high power Xingheng - high energy Xingheng - high energy Lantian Lantian Lantian Citic Guoan MGL Citic Guoan MGL Citic Guoan MGL Zhengke Panasonic AVERAGE-high energy

15 7.5

0.88 0.41

0.43 0.16

30 10 60 18 4.7 50 30 10 11 unavail.

1.0 0.37 1.8 0.6 0.14 1.95 1.1 0.47 unavail. unavail.

0.45 0.15 0.78 0.31 0.052 0.95 0.66 0.19 unavail. unavail.

111 100 123 111 124 97 104 81

249 241 286 215 333 201 173 198

510 527 unavail. unavail. unavail. unavail. unavail. unavail. 505 761

106

237

156

$586

Table 6: Comparison of Battery Types (with Assumptions) Results Cost ($) Mass (kg) Lifetime (yrs) Volume (L) Max Theoretical Power (kW) VRLA 75 26 3 10 6.2 Li-ion [30] 424 8 9 5 2.9

22

Recharging Safety Temperature effects Assumptions Specific Energy (Wh/kg) Energy Density (Wh/L) Power density (W/kg) Cost ($/kWh) Cycle Life

high moderate VRLA1 35 86 240 83 300

low high Li-ion 110 170 350 5051 800

1. Data for VRLA come from Chinese battery measurements and product brochures

Table 7: Performance characteristic of various cathode materials for Li-ion batteries [21] Cathode Material Energy density (Wh/kg) Cycle life (cycles) Price (US$/kg) LiCoO2 180 400 30 Li(Ni1/3Co1/3Mn1/3)O2 170 400 22 LiFePO4 130 1000 12

Zhengke Li-ion battery e-bike (anonymous source)

23